ERPs Reveal Atypical Neural Response During Empathy for Physical and Social Pain in ASD
Cora Mukerji1, Adam Naples1, Raphael Bernier2, Jennifer Foss-Feig1, Rachael Tillman1, Danielle Perszyk3, & James McPartland1
1Yale
Child Study Center, New Haven, CT
2Department
of Psychiatry,
3Northwestern
University, Evanston, IL
University of Washington, Seattle, WA
METHOD
BACKGROUND
Empathy is a core social ability affected in autism spectrum disorders (ASD)
manifest in:
•  Lower self-reported empathy and difficulties representing others’ mental states.
•  Atypical patterns of brain activation during emotion recognition and reduced
embodied empathy during physical pain.
Empathy for physical and social pain recruits distinct and overlapping neural
networks in typical development (TD):
•  Empathy for physical pain recruits affect-encoding brain regions (e.g., anterior
cingulate cortex, insula).
•  Empathy for social pain recruits mentalizing networks (e.g., dorsomedial
prefrontal cortex, precuneus).
•  Enhanced activation in affect-encoding regions during observed physical pain is
also seen in highly empathic individuals during empathy for social pain.
EXPERIMENTAL DESIGN
•  Three experimental manipulations:
•  Type of stimuli (Social/Physical)
•  Pain depicted in stimuli (Painful/Painless)
•  Task (Count/Rate)
+"
Rate or Count
+"
Image: 500 ms
250-500 ms
Distress?
1 = Not distressing
2 = Mildly distressing
3 = Moderately distressing
4 = Extremely distressing
Indefinite (Probe)
500 ms
Figure 2. Trial structure
ASD
N110
!100%
22 (3)
19-28
36 (14)
15-54
67 (18)
44-91
TD
21 (2)
18-25
38 (9)
27-56
--
--
•  The Empathy Quotient (EQ)
•  40-item measure of cognitive
empathy, emotional
reactivity, and social skills.
STIMULUS SET
•  60 dynamic videos and 60 static
images depicting hands in
painful/painless scenarios (15 per
condition).
Figure 1. Example static stimuli.
Static stimuli depict the last frame
of the preceding video.
200%
300%
400%
500%
600%
!100%
100%
P300
––– Physical Pain
- - - Physical Painless
––– Social Pain
- - - Social Painless
!1.5%
Latency (ms)
!2.5%
400%
500%
600%
P300
!1%
!1.5%
Figure 4b. Waveforms to observed physical and social
scenarios (painful and painless) in ASD.
Social
ASD
!1%
!1.5%
!2%
TD
!2.5%
Painful
90#
Figure 5. N110 amplitude to
observed painful and painless
scenarios in TD v. ASD (across
hemispheres).
350
400
450
500
+4#
70#
60#
•  Early emotional response to others’ pain may serve as an indicator for treatment
selection and a metric of outcome for social skills interventions in ASD.
50#
40#
+3#
+2#
+1#
0#
1#
2#
3#
Latency (ms)
N110 Amplitude (Physical - Social Pain)
Figure 6. P300 latency to
physical and social scenarios in
TD v. ASD (right hemisphere).
Figure 7. Correlation of SRS-A-SR
scores with a neural index of social pain
sensitivity at the N110 in ASD (left
hemisphere).
ACKNOWLEDGEMENTS
•  During cognitive stages of empathic processing, self-reported empathy was
associated with greater sensitivity to social pain at the P300 component in TD
and ASD.
•  The neural response to observed social pain is closely associated with empathic
functioning in both typical and atypical development.
80#
30#
Painless
•  SRS-A-SR scores in ASD correlate with N110 amplitude differentiation between
observed physical and social pain during the rating task [r=-.801, p=.030]:
•  Higher social function in ASD is associated with greater amplitude to social
relative to non-social pain in the left hemisphere.
IMPLICATIONS
100#
Physical
!0.5%
•  EQ scores correlate with the difference in P300 amplitude between observed
socially painful and painless scenarios in the right hemisphere across diagnostic
categories [r=.573, p=.007]:
•  Higher trait empathy is associated with greater amplitude to socially painful
versus painless scenarios.
•  ERPs indicated delayed neural processing of social actions and faster processing
of physical actions, as indexed by the P300 component, at cognitive stages of
empathic processing in ASD.
ASD
TD
ERP-BEHAVIORAL CORRELATIONS
•  EQ scores correlate with the difference in P300 amplitude between observed
physical and social pain in the right hemisphere across diagnostic categories
[r=-.641, p=.002]:
•  Higher trait empathy is associated with greater amplitude to social versus
physical pain.
•  During affective stages of empathic processing, enhanced sensitivity to social
pain at the N110 was associated with greater overall social functioning in ASD.
!2.5%
Figure 4a. Waveforms to observed physical and social
scenarios (painful and painless) in TD.
!3.5%
300%
!2%
!2%
!3%
200%
!0.5%
!1%
0%
0%
•  P300 Latency: Interaction of Diagnosis, Type, and Hemisphere [F(1,39)=9.425, p=.006]:
•  P300 latency to social scenarios longer than to physical scenarios in ASD
[t(6)=2.327, p=.059] but not in TD [t(13)=, p=.188] in the right hemisphere.
•  Right-hemisphere P300 latency to physical actions in TD>ASD [t(19)=2.791, p=.012].
•  ERPs revealed disruption of brain mechanisms regulating affective response to
observed pain/lack of pain, as indexed by the N110 component, in ASD.
N110
*
Physical Painful
100%
!0.5%
Amplitude (μV)
ASD
Amplitude (µV)
(μV)
Amplitude
SRS-A-SR Score
M (SD)
Range
0%
*
EQ Score
M (SD)
Range
0.5%
0%
0%
Age
M (SD)
Range
Social Painful
0.5%
ERP RESULTS
•  N110 Amplitude: Interaction between Diagnosis and Pain [F(1,19)=5.941, p=.025]:
•  N110 amplitude to painless scenarios in ASD>TD [t(19)=4.038, p=.001].
CONCLUSIONS
SRS-A-SR Score
TD
Table 1. Participant Demographics
Social Painless
750 ms
STATISTICAL ANALYSIS
•  Peak amplitude and latency were
analyzed using multivariate
repeated measures ANOVA.
•  4 within-subjects factors
•  Type (Social/Physical)
•  Task (Count/Rate)
•  Pain (Painful/Painless)
•  Hemisphere (Left/Right)
•  Between-subjects factor
•  Diagnosis (ASD/TD)
•  B i v a r i a t e c o r r e l a t i o n s w e r e
computed among amplitude and
latency difference scores and
behavioral scores.
RESULTS: FIGURES
PARTICIPANTS
•  14 TD male adults (2 left-handed)
•  7 male adults with ASD (1 left-handed)
Physical Painless
1500 ms
Video: 1500 ms
250-500 ms
METHOD
!
SELF-REPORT
BEHAVIORAL MEASURES
•  The Social Responsiveness Scale; Adult
Self-Report (SRS-A-SR)
•  6 5 - i t e m m e a s u r e o f s o c i a l
functioning in adults.
Figure 3. C3 and C4
recording sites
•  Four blocks, counterbalanced for sequence
•  60 trials per block (30 painful, 30 painless
trials randomized within block)
•  Each trial included a video and a static image
depicting the video’s final frame.
Neural markers of empathic response to social pain in ASD remain unexplored.
The current study investigates:
•  The temporal dynamics of empathy for physical and social pain in ASD versus TD.
•  Relations among neural responses to observed social pain, empathic traits, and
social function in ASD.
DATA ACQUISITION AND EXTRACTION
•  EEG recorded continuously at 250 Hz
•  Hydrocel Geodesic Sensor Net 128
•  Data segmented to static image (100 ms
pre-stimulus baseline, 550 ms poststimulus) and average-referenced
•  Peak amplitude and latency for the N110
(90-170 ms) and P300 (300-550 ms)
extracted at C3 and C4 sites
•  Task manipulated attention:
•  Count the bracelets on actors’ wrists (1-4)
•  Rate distress to observed pain (1-4)
Our previous work in TD adults revealed that autistic traits modulated eventrelated potential (ERP) markers of empathic processing for both physical and
social pain:
•  N110: a short-latency marker of empathic distress.
•  P300: an index of cognitive appraisal and stimulus categorization.
RESULTS: SUMMARY
The authors gratefully acknowledge the contributions of, Drs. Kevin Pelphrey and Linda Mayes and the Developmental
Electrophysiology Lab. This work was supported by NIMH K23MH086785 (JM), NARSAD Atherton Young Investigator
Award (JM), CTSA UL1 RR024139 (JM), and Yale College Dean’s Fellowship in Humanities and Social Sciences (CM).
•  Research in progress in our lab explores the modifiability of this response in
children and adults with ASD.
REFERENCES
1.  Baron-Cohen S, et al. J Autism Dev Disord. 2004;34(2):163-175.
2.  Gremiel E, et al. Neuroimage. 2010;49(1):1055-1065
3.  Fan Y, et al. Neuropsychologia. 2008;46:160-173.
4.  Cheng Y, et al. Neuroimage. 2008;40(4):1833-1840.
5.  Masten C, et al. Neuroimage. 2011;55(1):381-388.